|
 |
|
Selected Models
|
- 2-Compartment model
- with K1/k2 constraint
- Patlak, Logan plots
- Time-weighted integral
method
- Simplified reference model
- Ichise’s reference models
- Logan’s non-invasive model
- Glucose autoradiography
- Fractal analysis
- Fourier analysis
|
|
|
|
|
18F Autoradiography (Rat)
|
|
| Pixel-wise Modeling Tool (PXMOD) |
- Calculation of models in each image pixel
- Models include perfusion, distribution volume,
glucose turnover, binding potential, and more
- Joint exploration of time-activity curve and model results
- Reference models without the need for blood sampling
|
 Brochure |
[ Case Gallery - Neurology] |
|
|
| Absolute Function from Nuclear Medicine Data. Although called „functional“ images, reconstructed emission tomography images just represent the average tracer activity during the acquisition. This snapshot depends on several factors such as the injected dose, time after injection, and the tracer uptake parameters. Successful tracers have the property, that one main parameter is emphasized, while the other factors are less relevant for the image. Still, the images are only „functionally weighted“, but not in absolute units such as tissue perfusion in [ml/min/ml tissue] and can not truly be compared across studies. However, if the tracer uptake is recorded with a suitable acquisition protocol, a model-based analysis can be performed in each image pixel. It results in a functional image showing one isolated tissue parameter. With the present pixel-wise modeling tool such functional images can efficiently calculated with over 20 methodologies. |
|
| Supports Comprehensive and Intuitive Pixel-wise Data Analysis. After loading blood and dynamic image data, a pipeline of three processing stages is performed. First, externally sampled blood data undergoes some corrections such that afterwards it represents blood activity at the site of the analysis. Second, a model pre-processing step is performed which calculates quantities relevant for model evaluation in all pixels. Third, the selected model is applied to the time-activity vector in each pixel, and for each result parameter a functional image is formed. The results can be explored with an inspector which shows the time vector in parallel with the result parameters, and saved in one of the many formats. An additional feature supports further results evaluation: if matched anatomical images are available for the patient they can directly be loaded and fused with the functional maps. In order to optimize reproducibility and throughput, standard processing protocols can be defined for certain study types. |
|
| Selected Applications |
- rCBF from water PET studies
- MrGIu from FDG PET studies
- Binding potential and k3/k4 from neuro-receptor studies
- 2-Compartment model parameters with K1/k2 constraint
|
- Distribution volume
- Correlation coefficients
- Regression analysis
|
|
|
Products 
Pixel-wise Modeling]
